Paper to be presented at

DRUID15, Rome, June 15-17, 2015

(Coorganized with LUISS)

ACQUIRED INVENTORS? PRODUCTIVITY AFTER HORIZONTAL

ACQUISITION: MANAGING THE R&D INTEGRATION PROCESSMassimo ColomboPolitecnio di Milano

Department of Management, Economics and Industrial Engineerimassimo.colombo@polimi.it

Solon Moreira

University of Bath,

sm.ino@cbs.dk

Larissa RabbiosiCopenhagen Business School

Department of International Economics and Managementlr.smg@cbs.dk

AbstractEffective integration of the R&D functions of the acquired and acquiring firms is essential for knowledge recombinationafter acquisition. However, prior research suggests that the post-acquisition integration process often damages theinventive labor force. We argue that an examination of the multifaceted nature of the integration process furtherenhances our understanding of which conditions will be more or less detrimental for corporate inventors. We focus onR&D teams which are the immediate organizational context in which inventors operate and drawing on insights fromlearning theory and evolutionary economics we posit and find that the reorganization of R&D teams after acquisitionharms acquired inventors? innovative performance. Though, the implementation of other integration decisions canmitigate or aggravate this negative effect.

Jelcodes:O31,L19

 

1

 

ACQUIRED INVENTORS’ PRODUCTIVITY AFTER HORIZONTAL ACQUISITION:

MANAGING THE R&D INTEGRATION PROCESS

ABSTRACT

Effective integration of the R&D functions of the acquired and acquiring firms is essential for

knowledge recombination after acquisition. However, prior research suggests that the post-

acquisition integration process often damages the inventive labor force. We argue that an

examination of the multifaceted nature of the integration process further enhances our

understanding of which conditions will be more or less detrimental for corporate inventors. We

focus on R&D teams which are the immediate organizational context in which inventors operate

and drawing on insights from learning theory and evolutionary economics we posit and find that

the reorganization of R&D teams after acquisition harms acquired inventors’ innovative

performance. Though, the implementation of other integration decisions can mitigate or

aggravate this negative effect.

INTRODUCTION

Horizontal acquisitions—i.e., acquisitions of firms that operate in the same industry as the

acquiring firm—are an important mechanism firms use to grow and increase their competitive

advantage. A consolidated stream of studies has examined the economic and financial

performance of horizontal acquisitions (for a review, see Haleblian et al., 2009). This literature

shows that the realization of synergies that could arise from combining resources and capabilities

of the acquiring and acquired firms strongly depends on the post-acquisition integration process

(e.g., Capron, 1999; Larsson and Finkelstein, 1999; Ranft and Lord, 2002; Zollo and Singh,

2004). Integration involves “changes in the functional activity arrangements, organizational

structures and systems, and cultures of combining organizations to facilitate their consolidation

into a functioning whole” (Pablo, 1994: 806), and poses serious challenges to managers in the

newly created firm. Indeed, failures and poor performance of horizontal acquisitions are often

associated to an ineffective implementation of the integration process (e.g., Jemison and Sitkin,

 

2

 

1986; Shrivastava, 1986; Haspeslagh and Jemison, 1991; Larsson and Finkelstein, 1999). The

attention toward the integration process becomes paramount also in relation to an increasing

consideration of scholarly work on the innovation impact of horizontal acquisitions (e.g.,

Capron, 1999; Ahuja and Katila, 2001; Cassiman et al., 2005). Research shows that the

interdependence between firms’ technology strategy and the post-acquisition integration process

explains firms’ ability to be innovative and reap synergistic technological gains after the

acquisition (Graebner, 2004; Puranam, Singh, and Zollo, 2006; Colombo and Rabbiosi, 2014).

In line with an evolving stream that calls for exploring the micro-foundations of firm-

level studies (Felin and Foss, 2005; Felin and Hesterly, 2007), parallel work has started

examining the relationship between horizontal acquisitions, integration and innovation at the

individual level of analysis (Paruchuri, Nerkar, and Hambrick, 2006; Kapoor and Lim, 2007;

Paruchuri and Eisenman, 2012). The focus is on a specific discrete integration decision—

structural integration—consisting in the absorption of the acquired operations within the

organization of the acquiring firm as opposed to keeping the acquired operations as a separate

subsidiary. The main finding is that structural integration negatively affects the post-acquisition

innovation productivity of acquired inventors (Puranam et al., 2006). These studies have

enriched our understanding of the effects of integration on innovative performance at the

individual level in the very important context of the acquisitions of small-technology-based

firms. However, horizontal acquisitions often involve larger firms where post-acquisition

integration hinges on a broader set of different decisions which are likely to affect the innovative

behavior of acquired individuals. These decisions aim at physical integration like changes of the

organization structure (structural integration is one possible decision, others are the divestiture of

assets and the elimination of redundancies), technological integration such as changes of the

 

3

 

research scope or reallocation of R&D resources, and managerial integration like changes of

responsibilities and decision-making power (Shrivastava, 1986; Haspeslagh and Jemison, 1991;

Pablo, 1994; Capron, Mitchell, and Swaminathan, 2001). Post-acquisition integration is complex,

nonetheless, how these complexity influences the innovation performance of acquired

individuals’ behavior and productivity is still unclear.

We contribute to addressing this research gap by looking at how different integration

decisions, implemented in isolation or in combination, impact the post-acquisition innovation

productivity of individual acquired inventors. First, we consider integration through the

reorganization of R&D teams in the period that immediately follows the acquisition. More

precisely, integration occurs if inventors of the acquiring and acquired firms are transferred

across the two firms and organized into new teams. R&D teams are the immediate organizational

context in which inventors operate. Recombining R&D teams is a powerful decision to facilitate

the transfer of knowledge embedded in individual inventors, foster inter-personal communication

and coordination, and integrate the two firms’ technological capabilities (Ranft and Lord, 2000;

Ranft and Lord, 2002). However, the reorganization of teams puts pressure on individual

inventors to develop new or modify existing working routines (Nelson and Winter, 1982;

Szulanski, 1996; Edmondson, 2002). We posit that the benefits of knowledge recombination on

inventor productivity stemming from the reorganization of R&D teams are overcome by the

costs the same reorganization raises challenging the collective learning process occurring at the

team level. Second, given the complexity of the integration process we argue that hardly an

integration decision works in isolation while it is likely to be contingent on situation created by

other decisions. Accordingly, we focus on three additional post-acquisition integration decisions

occurring at the physical-, technological-, and managerial level of integration which we expect to

 

4

 

indirectly influence inventors’ collective learning. These decisions are the divestiture of acquired

R&D laboratories, the launch of R&D programs in technological fields new to the acquired firm,

and the replacement of the acquired R&D top manager. We posit that the negative effect of R&D

team reorganization is aggravated when acquired R&D laboratories are closed and projects in

technological fields new to the acquired inventors are launched after the acquisition. On the other

hand, we suggest that favoring unlearning of existing routines the replacement of the acquired

R&D top manager mitigates the negative effects on inventor productivity induced by R&D team

reorganization.

Our empirical analysis is based on the patenting activity of 3,693 acquired inventors who

continue to work within the new firm after the acquisition (Paruchuri et al., 2006; Kapoor and

Lim, 2007) and 25 horizontal acquisitions of sizable firms in medium- and high-tech industries.

In each acquisition, the R&D function is a component of the acquired firm’s assets and the R&D

operations of acquiring and acquired firms are in the same broadly defined technological area.

We construct our sample implementing the coarsened exact matching (CEM) technique and

empirically test our arguments applying a difference-in-differences setup in a longitudinal data

setting.

Our study contributes to extend the work on post-acquisitions integration at the

individual-level of analysis. Scholars tend to confirm that the decline in the patenting activity

after acquisition observed at the firm-level (Hitt et al., 1991) is associated with a decline in

patenting at the individual level (Kapoor and Lim, 2007) and the negative effect on inventor

productivity is worsened by structural integration when small technology-based firms are

acquired (Paruchuri et al., 2006). In the context of horizontal acquisitions of sizable firms, our

findings suggest that interdependencies between different integration decisions can help

 

5

 

explaining the differential effects of post-acquisition integration on inventors. Moreover, our

work adds value to the general literature on post-acquisition integration studying how this process

occurs through the reorganization of R&D teams that are the organizational context where

knowledge recombination is more likely to take place. Finally, we theoretically argue and

empirically corroborate that through further integration decisions firms can mitigate or aggravate

the negative effect of R&D team reorganization on inventors’ innovation productivity. Accordingly,

we contribute to capture the multifaceted and complex nature of post-acquisition integration.

POST-ACQUISITION INTEGRATION PROCESS

Value creation as reflected in the combination’s potential of synergistic assets and resources that

have initially motivated the acquisition are often not realized. Research on post-acquisition

implementation highlights that this is often the case because of the poor and ineffective process

through which resources are integrated (e.g., Jemison and Sitkin, 1986; Datta, 1991; Pablo,

1994). When two previously separated firms come together under the same corporate structure,

the realization of synergies depends on integration through the rationalization, reconfiguration,

and coordination of resources and assets of the acquiring and acquired firms (Haspeslagh and

Jemison, 1991). Integration can be of different intensity (from low to high) and involve different

decisions. Post-acquisition integration can be low and consists in minor changes like the

standardization of basic processes and systems, or more widespread including changing in

physical and knowledge-based resources like the consolidation of product lines and R&D

projects, and the redeployment of resources (e.g., Shrivastava, 1986; Napier, 1989; Pablo, 1994;

Capron, Dussauge, and Mitchell, 1998; Larsson and Finkelstein, 1999; Capron et al., 2001; Ranft

and Lord, 2002; Graebner, 2004; Zollo and Singh, 2004).

 

6

 

A broad set of studies at the firm-level of analysis point to several important post-

acquisition integration decisions and discuss their effects on innovation. For instance, the

redeployment of resources of the acquired and acquiring firms enhances product features (e.g.,

product cost, product quality) (Capron and Hulland, 1999) and improves the overall

organizational innovation capabilities (Capron, 1999). Graebner (2004) highlights that if the

acquired top management remains in place after the acquisition the success of the integration of

acquired and acquiring firms’ resources is more likely. The implementation of rich

communication systems like “cross-fertilized” teams of managers and formal and informal

events facilitating face-to-face interactions is also essential to establish a favorable climate

between acquired and acquiring employees and facilitate two-way flows of knowledge and

integration of technologies and capabilities (Ranft and Lord, 2002).

At the individual level of analysis, the understanding of the effect of integration on

innovative performance is limited to the role played by post-acquisition structural integration.

Structural integration tends to decrease individual inventor productivity because it reduces the

autonomy of employees, harms their motivations, increases uncertainty about career prospects

and creates a sense of dislocation and anxiety (Paruchuri et al., 2006). While structural

separation can be considered a proper strategic option in the case of acquisitions of small-

technology-based firms, when larger firms are acquired structural integration is more likely to

take place and precede the implementation of a broader set of additional integration decisions

(Pablo, 1994). In the attempt to understand the effect of organizational integration on inventor

productivity, in this study we consider four integration decisions which are presented in the next

sections.

A look at the complexity of R&D post-acquisition integration

 

7

 

We expect that the integration of physical, technological, and managerial resources provides new

opportunities for knowledge combination and the conditions for different and better use of R&D

resources which, ultimately, can enhance individual inventors’ productivity. However, as we discuss

later in the paper, integration decisions taken to combine two previously separated R&D functions

also generate costs which, ultimately, can abate the potential positive effect of integration.

Physical integration of resources and assets involves the decision to eliminate common

inputs (e.g., reduction of R&D personnel, closure of R&D laboratories) (Shrivastava, 1986;

Bowman and Singh, 1993; Capron et al., 2001; Colombo and Rabbiosi, 2014). A situation

typical of horizontal acquisition is that assets of the acquiring and acquired firms in part overlap

and in part are complementary and mutually exclusive. This relatedness forms the basis for

mutual understanding and realization of potential synergies (Ahuja and Katila, 2001). However,

the same relatedness gives raise to redundancies and duplications. The rationalization of some

R&D activities and the redeployment of others might be necessary to achieve synergy-based

benefits from the acquisition. Technological integration involves decisions affecting the scale of

typical R&D projects, the focus on specific technological fields, the specialization or

diversification of R&D tasks, the re-design of research teams. After the acquisition, research

projects that were not feasible before become feasible through the exchange of resources. Thus,

technological integration allows for new combination of different R&D inputs to potentially

realize knowledge and outputs that did not exist before or achieve efficiencies that could not be

achieved previously or only at prohibitive costs (Ahuja and Katila, 2001; Cassiman et al., 2005).

Finally, as part of the integration process, managerial integration involves a complex combination

of decisions related to the selection and transfer of managers after acquisition (Shrivastava,

1986). The turnover of the acquired top management is often seen in relation to the integration

 

8

 

process because managers represent repositories of potentially valuable knowledge that can be

exploited during integration (e.g., Graebner, 2004; Cording, Christmann, and King, 2008).

We capture the complexity of post-acquisition R&D integration analyzing four

integration decisions. To take into account physical integration, we study the divestiture of

acquired R&D laboratories after the acquisition while we capture managerial integration with the

decision to maintain or replace the acquired R&D top manager. In the case of technological

integration, we focus on two decisions that are the reorganization of the R&D teams and the

launch of R&D programs that are new relating to the technological fields in which the acquired

firm  had previously developed distinctive capabilities. These four decisions are all crucial for

R&D integration. However, as we argue in the next section, the reorganization of R&D teams is

paramount while we expect the other decisions to amplify or reduce the direct effect of R&D

team reorganization on inventor productivity.

R&D post-acquisition integration: the reorganization of R&D teams

The knowledge-based view of the firm plays a central role in explaining how knowledge

recombination can occur and synergies be realized after acquisition. Taking its intellectual roots

in the behavioral theory of the firm (Cyert and March, 1963) and evolutionary economics

(Nelson and Winter, 1982), the knowledge-based view identifies firms as generators, repository,

and integrators of knowledge (e.g., Kogut and Zander, 1992; Grant, 1996). Post-acquisition

resource recombination’s potential can be seen as result of a knowledge-based process in which

acquiring and acquired firms learn how to transfer, teach and utilize each other’ existing tangible

and intangible resources and create new knowledge (e.g., Kogut and Zander, 1993). This

learning process is facilitated by the superior coordination provided by having people, practices

and intellectual property embedded in the same organizational entity (Kogut and Zander, 1993;

 

9

 

Grant, 1996; Kogut and Zander, 1996) and by the fact that acquiring and acquired firms share

related prior knowledge and contextual understanding (Cohen and Levinthal, 1990).

Organization’s knowledge is encapsulated in individual skills and organizational

routines: skills refer to knowledge and abilities held by employees (i.e., firm resources) while

routines can be described as recurrent interaction patterns (Nelson and Winter, 1982). Routines

involve multiple resources linked by interactions and are collective phenomena. Specifically,

routines capture the joint use of knowledge held by individuals organized in activities (e.g.,

teams), which, however, is more than the sum of the knowledge held by the single individuals

(Becker, 2004). Routines are a result of path-dependent actions taken over time and bound by the

firm’s existing technological base and knowledge (e.g., Dosi, 1982; Dierickx and Cool, 1989).

Research outputs such as patented innovations reside in the interlinked and mutually

reinforcing skills of individuals with the organizational routines connecting these individuals

(Nelson and Winter, 1982). Knowledge recombination relies on an intricate, path-dependent

context in which inventors search locally and (implicitly) apply routinized tasks (e.g. information

filters, communication and problem-solving strategies) to give rise to innovation, and requires

individual to communicate and coordinate to create new routines, thereby participating in a

collective process (Nelson and Winter, 1982; Henderson and Clark, 1990). Given the collective

nature of knowledge recombination, we identify the decision to reorganize R&D teams in the

period that immediately follows the acquisition as a key decision the newly created firm can

implement after acquisition to integrate the previously separate R&D functions and facilitate

synergistic value creation. More precisely, integration occurs if inventors of the acquiring and

acquired firms are transferred across the two firms and organized into new teams.

 

10

 

How that team reorganization is such an important integration decision for individual

inventors? We expect an inventor’s ability to unlock the post-acquisition innovative potential

stemming from the realization of synergies to depend on the capacity of his/her R&D team to

encourage the coordinated exchange of knowledge, materials, and reciprocal inputs and to work

interactively to complete research tasks. Since we are interested in the effect of post-acquisition

integration at the individual-level of analysis, the focus on R&D team reorganization allows to

observe the effect of the integration process in the context in which dialogue, discussion,

experimentation, and reflection between knowledge-workers take place (Senge, 1990). Teams

hold together set of individuals who share responsibilities and assumptions, acquire, share, and

combine knowledge, and carry out activities that leave outcomes (Alderfer, 1987; Hackman,

1987; Argote, Gruenfeld, and Naquin, 2001). Teams learn though their team members’ ability to

share information, seek for feedback and ask questions, talk about errors and reflect on results

(Edmondson, 1999). At the same time, team learning also benefits the members who compose

the team in a way that learning in teams helps developing the knowledge and skills of team

members (Edmondson and Nembhard, 2009). Post-acquisition R&D team reorganization

facilitates each inventor to access not only self-contained knowledge and skills from her

organization’s knowledge base but also inventors’ skills and the organizational knowledge base

of the other firm. As a result, the team and its members become embedded in multiple, non-

redundant resources which transfer and use can manifest through research output like patents.

THEORY AND HYPOTHESES

The effect of R&D team reorganization on inventors’ productivity after acquisition

The reorganization of teams is a powerful integration decision because it creates the necessary

conditions for different inventors to engage in face-to-face communication and share their

 

11

 

knowledge (Ranft and Lord, 2000; Ranft and Lord, 2002). Since the acquired and acquiring

firms share resources that are similar or complementary—i.e. they are related—and relevant

within each other’s context the recombination of knowledge will be eased (Haspeslagh and

Jemison, 1991; Ahuja and Katila, 2001). In other words, shared industry experiences and related

technical and knowledge bases of inventors of the acquiring and acquired firms provide a

contextual understanding that facilitates the communication and sharing of inventors’ specific

knowledge among team members (Cohen and Levinthal, 1990). However, although inventors

may share organized knowledge structures and routines stemming from the relatedness in their

technical and scientific skills, the reorganization of R&D teams can also damage team-level

collective learning.

With team reorganization team members are reshuffled to include inventors of both the

acquired and acquiring firms. Thus, reorganized inventors face a pressure to develop new or alter

existing routines—or shared mental models (Edmondson, 2003).1 With post-acquisition

integration existing routines of the acquiring (acquired) inventors need to come together with

those routines of the acquired (acquiring) inventors. Knowledge recombination requires intense

team members’ interdependence—“the extent to which team members cooperate and work

interactively to complete tasks” (Stewart and Barrick, 2000: 137). Interactions develop strong

ties, members’ cohesion, team identity and coordination that facilitate informal learning among

team-members (Mudrack, 1989; Manz and Newstrom, 1990). However, coordination routines

will suffer from the reorganization of teams: while keeping team members together facilitates

communication and coordination and enables the development of routines that support new task

                                                            1 The team literature describes mental models as organized knowledge structures that allow individuals to describe,

explain, and predict actions in their environment (Klimoski and Mohammed, 1994). When individuals share mental

models they select actions that are consistent and coordinated with those of the other team members (Mathieu et al.,

2000). Team research shows that the functioning and performance of teams are related to shared mental models in

teams (for a review, see Mohammed, Ferzandi, and Hamilton, 2010).

 

12

 

learning (e.g., Levine, Richard, and Moreland, 1999) the reorganization of teams dismantles

team routines and negatively affects team performance (Lewis, Lange, and Gillis, 2005).

Moreover, the way how team inputs are transformed into outputs are influenced by team routines

(Kraiger and Wenzel, 1997). In particular, team members who share mental models work toward

common objectives, easily coordinate their actions, and exhibit effective team processes, which

subsequent positively affect team performance (Mathieu et al., 2000; Rentsch and Klimoski,

2001). However, facing new unfamiliar work environments team members cannot apply existing

routines and need to learn new ones to succeed (Rico et al., 2008).

Although R&D team reorganization fosters communication and coordination among

acquired and acquiring inventors, these inventors may have difficulty in interact with each

other’s given their different routines. They may also underestimate the amount of extra learning

necessary to effectively take on the new shared mental models, indeed the convergence on new

routines develops out of salient shared experiences (Nelson and Winter, 1982). Therefore, we

posit that integration through team reorganization will prove frustrating, demobilizing and

inefficient for inventors who will have to invest more effort in learning and aligning with the

dominant (or new) routines of the team.

Hypothesis 1: The patenting performance of acquired inventors will decrease more after

acquisition if post-acquisition integration involves the reorganization of R&D teams of

the acquired and acquiring firms.

Moderating effects: looking at the complexity of post-acquisition R&D integration

The first integration decision we expect to moderate the effect of R&D team reorganization on

inventors’ productivity is the divestiture of acquired R&D laboratories which captures physical

integration. Horizontal acquisitions are often characterized by overlapping activities and

 

13

 

redundancies which gives the opportunity to downsizing and redeploy the remaining resources

into a more efficient recombination (Capron, 1999; Capron et al., 2001). We suggest that when

the reorganization of R&D teams is concurrent with the divestiture of acquired R&D

laboratories, the negative effect of team reorganization on the acquired inventors’ productivity

will be worsen by the sense of disruption acquired inventors will experience.

The achievement of synergistic benefits after the acquisition not only requires integration

but also to maintain morale and gain the commitment of people to the new goals (Shrivastava,

1986; Ranft and Lord, 2000). Closing down R&D laboratories has a direct impact on the morale

of the retained employees. Although people might feel relief about their retained jobs, they also

become emotionally aroused about their negative expectations regarding the future. Specifically,

resource redeployment after rationalization is often accompanied by geographical relocation or

changes of the internal affiliation. Accordingly, after divestitures employees will feel uncertain

about their status and career and concerned of losing their identity (e.g., Walsh, 1989; Paruchuri

et al., 2006). In addition, divestiture of acquired R&D assets is also likely to exacerbate the

feeling of submission and inferiority of acquired employees (e.g., Hambrick and Cannella,

1993). Asset divestiture builds up sense of disruption, resentment, and hostility that aggravates

the working conditions of acquired inventors who will express their uncertainty and feeling of

dislocation in subversive behavior and reduced productivity (Shrivastava, 1986; Ernst and Vitt,

2000; Ranft and Lord, 2000). Thus, we expect the loss of motivation and disruption induced by

rationalization to worsen the negative effect of team reorganization of inventor productivity.

Hypothesis 2: The closure of acquired R&D laboratories aggravates the negative effect

of post-acquisition R&D team reorganization on acquired inventors’ patenting

performance after the acquisition.

 

14

 

Building on the literature of team mental models we distinguish between task- and team-related

routines to examine the moderating effect of the two remaining integration decisions. Team

researchers proposed the existence of multiple and simultaneous mental models (Cannon-

Bowers, Salas, and Converse, 1993), which can be reflected into two major conceptually distinct

types, one related to task work and one related to team work (Mathieu et al., 2000). The former

refers to the ability of team members to understand the technology and equipment with which

they are interacting and their sharing of knowledge about how the task is realized (e.g.,

knowledge of immunology held by team developing immunotherapy drugs; knowledge of

photolithographic aligners held by team developing semiconductor devices). We refer to these

aspects as task-related routines. On the other hand, team work mental models types relate with

team members sharing knowledge about team interactions (e.g., roles and responsibilities in the

team, communication channels, information sources) and team models (e.g., knowledge about

what to expect from team members). We refer to these aspects as team-related routines.

We posit that when the reorganization of R&D teams is concurrent with the launch of

R&D projects in technological fields new to the acquired firm, the negative effect of team

reorganization on the acquired inventors’ productivity will be worsen by the supplementary

challenges the exposure to research tasks in new technological fields induces on acquired

inventors’ team-related routines. This expectation is consistent with the fact that R&D programs

in new technological fields require existing inventors’ routines to deal with the use and

understanding of new equipment and technology which change team members’ tasks, and make

the previously developed routines not sufficient anymore (Edmondson, 2003). It is well

recognized that new technologies lead to changing established task-related routines at different

levels—organizational, team, and individual—and that this process is not without a struggle

 

15

 

(e.g., Orlikowski, 1993; Edmondson, Bohmer, and Pisano, 2001). Moreover, previous work

suggests that teams whose members maintain tasks in the same domain are more likely to

develop routines relevant to the task domain—a critical factor for learning (Lewis et al., 2005).

Task knowledge is also necessary for the recognition of members’ task-relevant expertise that is

found to improve a team’s task performance (Stasser, Stewart, and Wittenbaum, 1995). Finally,

team members’ differences in task-related routines result in ineffective team processes and,

therefore, in a reduction of team members’ performance (Mathieu et al., 2000). Based on this

reasoning, we suggest that:

Hypothesis 3: The launch of projects in technological fields new to the acquired inventors

aggravates the negative effect of post-acquisition R&D team reorganization on acquired

inventors’ patenting performance after the acquisition.

The idea that top management turnover can be used to force the unlearning of old routines and

overcome the inertia of organizational change finds its roots in pioneering work in strategic

management (e.g., Nystrom and Starbuck, 1984; Tushman and Romanelli, 1985; Wiersema and

Bantel, 1993). Top managers tend to stick to existing cognitive structure, practices, and

procedures which inhibit the exploration of alternative routines, rules, goals, etc. (Wiersema and

Bantel, 1993). Because top managers can directly influence and shape organizational norms and

values, top managers’ replacement disconfirms past programs, interrupts ongoing activities and

interrelationships and is a way to break inertial factors which operate to maintain the status quo

(Tushman and Romanelli, 1985). As the replacement of top managers is needed to overcome

inertia and accomplish the adaptation to changing contexts (Wiersema and Bantel, 1993), we

suggest the replacement of acquired R&D top managers as an important managerial integration

 

16

 

decision to overcome acquired inventors’ inertia to change and enforce the adoption of new

team-related routines.

Team-related routines rely on sharing of knowledge about how inputs are transformed

into outputs by team members. These routines stem from interaction patterns involving

individual knowledge about team interactions and team expected behaviors. Examples are team

members’ shared understanding of the amount and quality of knowledge, the information sources

and flows necessary to satisfy the communication dimension, the coordination of actions, roles

and responsibilities directed to predict the nature of team interactions and strategy, the attributes

directed at tailoring what to expect in terms of team cooperation. Given their “how to do” nature,

it is reasonable to expect that team-related routines of team members will share similarities with

managers’ routines. In confirmation of this, Klein and Posey (1986) find that supervision tends to

improve if managers and teams share the same team-related routines. Our expectation is

supported also by the fact that team researchers suggest organizational culture and environment

in which teams operate to influence team mental models about how to encode, store, retrieve,

and communicate information (Kraiger and Wenzel, 1997; Arrow, McGrath, and Berdahl, 2000).

However, also managers’ routines are influenced by their social and cultural environment (Huff,

1982). In addition, managers directly participate to social interaction with other individuals

within the organization and this social interaction results in shared organized knowledge

structures among the individuals involved—e.g., managers and team members (Klimoski and

Mohammed, 1994). Finally, organizational activities are driven by routines developed in a given

environmental configuration and adaptations to fit the external changes are dominated by top

managers’ idea and cognitive schema (e.g., Nystrom and Starbuck, 1984). Therefore, managers

 

17

 

play a unique role to create, shape and reinforce organizational routines which in general are

implicitly internalized by the other individuals in the firm.

To the extent that manager’s cognitions and routines are similar to team-related routines

shared by inventors, responses of inventors to organizational stimuli that are not aligned with

meeting a target will not be organizationally sanctioned. However, routines that are not

associated with failures will continue to be perceived acceptable and persist in the organization

(Cyert and March, 1963). Accordingly, the likelihood that acquired inventors will adapt faster to

the new team-related routines will depend on the organization’s ability to associate acquired

existing routines with failure. This is difficult to achieve when acquired R&D top managers are

retained given their shared understanding of acquired inventors’ team-related routines. Only by

removing the acquired R&D top manager the organization can clearly signal the separation from

the acquired team-related routines and elicit the acquired inventors to adjust or substitute their

team-related routines with the new dominant ones. In other words, the replacement of the

acquired R&D top manager will allow for the unlearning of the old team-related routines

facilitating the discovery of their inadequacies in the new research context.

Hypothesis 4: The replacement of the acquired R&D top manager alleviates the negative

effect of post-acquisition R&D team reorganization on acquired inventors’ patenting

performance after the acquisition.

METHOD

The test of the effect of post-acquisition R&D team reorganization on inventor performance

requires a methodology that takes into account selection effects and unobserved heterogeneity.

First, we need to account for the fact that characteristics related to the inventors in the pre-

acquisition period could trigger the decision to reorganize teams. For example, if inventors with

 

18

 

inherent low patenting performance are more likely to be in firms that reorganize R&D team,

then their post-acquisition performance vis-à-vis to inventors that did not go through a process of

reorganization would be naturally lower. Therefore, a cross-section estimation of the post-

acquisition period would not be an appropriate alternative. Second, inventors exhibit a life cycle

in their innovation productivity (Levin and Stephan, 1991; Kapoor and Kwanghui, 2007), with

the rate at which they produce innovations varying with the experience that they accumulate.

Therefore, it is not enough to estimate the effect of R&D team reorganization solely based on pre

and post individuals’ performance because this approach would provide no indication of

abnormal deviations from what is expected to be observed for an individual within a similar

trajectory. Finally, previous work shows the negative effect of acquisitions on inventor

productivity (e.g., Paruchuri et al., 2006; Kapoor and Kwanghui, 2007). Accordingly, to estimate

the effect of R&D team reorganization on inventor’s patenting performance, we first need to

account for the expected decline in inventors’ productivity triggered by the acquisition itself. Not

accounting for this effect could lead to an overestimation of the negative effect of post-

acquisition integration through R&D team reorganization.

In the attempt to overcome these endogeneity issues, we apply a difference-in-differences

setup in which we compare the patenting performance of acquired inventors involved in R&D

team after acquisition (treatment group) with inventors which were not exposed to R&D team

reorganization after acquisition (control group). For both groups we observe the longitudinal

changes in their performance based on the pre- and post-acquisition periods. In order to

implement this empirical strategy, we rely on the Coarsened Exact Matching (CEM) technique to

estimate the ATT (average treatment effect on the treated) concerning inventors patenting

performance. This matching technique is based on the improvement of the balance between

 

19

 

treatment and control groups making their joint distributions in terms of the covariates more

similar (Iacus, King, and Porro, 2011; 2012).

Sample and data

The sample consists of 3,693 acquired inventors working in 25 firms resulting from horizontal

acquisitions undertaken in medium and high-tech industry by large firms headquartered in the

European Union (EU). We identify these horizontal acquisitions based on data collected as part

of a research project promoted by the DG Research of the European Commission.2 Using the

1993 EU Market Share Matrix (MSM), the project identified 59 acquiring leading firms—the

five largest EU producers in EU 3-digit medium and high-tech industry in that year—

headquartered in EU. 31 firms out of 59 (response rate of approximately 52%) agreed to

participate in the study based on a multiple-case design and provided detailed information about

31 acquisitions occurred during the 1987-2001 period. Some acquisitions were domestic, other

cross-border and involved firms located in other EU countries as well as in North America. The

choice of selecting acquisitions from the DG research project was dictated by the fact that is

usually very difficult to have access to detailed information about post-acquisition integration

processes. The DG project data were collected during face-to-face interviews with who were in

charge of or actively participated in the implementation of the acquisition process (in most cases

the Vice-President for strategy or corporate development and the Vice-President for R&D or the

Chief Technology Officer). This research design allowed for collecting specific data about post-

acquisition decisions aimed at integrating the R&D functions of the acquired and acquiring

firms. We use this information to operationalize the integration decisions examined in our study.

                                                            2 These case studies were conducted within the FP5 project “Mergers and Acquisitions and Science and Technology

Policy” funded by the European Commission, DG Research (Contract No. ERBHPV2-CT-1999-13). 

 

20

 

The innovation productivity of each acquired inventor is defined using the Patent

Network Dataverse database that builds on the United States Patent and Trademark Office

(USPTO) and includes information regarding the identification of disambiguated names of

individual inventors registered at the U.S. utility patents for the period 1975−2010 (Li et al.,

2014). Focusing on a single large patent office ensures data consistency and avoids potential

noise created by cross-offices differences in terms of procedures and practices regarding the way

patent applications are managed. Moreover, the U.S. market represents one of the world’s largest

markets for global firms in medium- and high-tech industry like those included in the DG project

research. The following steps describe the construction of our dataset. First, based on the

identification of the company name we identified all patents filed at the USPTO by the 31

acquired firms in the 5 years prior the acquisition. We removed from the sample five firms with

no patenting activity. Second, in some cases the acquisition involved only specific businesses of

the acquired firm. Accordingly, we scrutinize the patent data of the remaining 26 acquired firms

to retain only information pertaining to the business units and their inventors directly involved in

the acquisition3. However, in one case the complex structure of the assets of both the acquired

and acquiring firms involved in the acquisition did not allow us to uniquely identify businesses

and inventors that undoubtedly were part of the acquisition. We removed this case from our

sample and remained with 25 deals involving 4,785 unique inventors that actively patented

during the pre-acquisition period—i.e., active inventors. Third, we accounted for inventors

mobility across firms (Singh and Agrawal, 2011; Younge, Tong, and Fleming, 2014).

                                                            3 For example, in the acquisition of Westinghouse by British Nuclear Fuels Limited (BNFL) only the nuclear

business of Westinghouse was acquired. To identify the inventors that were part of the acquisition, we scrutinized

the abstracts and content of all patents filled by Westinghouse and select only those inventors involved in the

nuclear business. Similarly, in the acquisition of CytoMed by UCB the deal involved activities of CytoMed in the

fields of allergy, asthma and central nervous system. Using the same approach, we selected only CytoMed inventors

related to the patents that were part of the deal.

 

21

 

Specifically, we removed inventors who had patented in a different firm within the five years

before and after the acquisition was concluded—the assignee name differs from the name of the

acquired firm. After removing these inventors, the final sample consists of 3,978 active acquired

inventors.

Measures

Consistent with prior research on inventors patenting performance (e.g., Paruchuri et al., 2006;

Kapoor and Kwanghui, 2007; Singh and Agrawal, 2011), we measure our dependent variable

patenting performance as total number of granted patents filed by an inventor with the acquired

firm during the pre- and post- acquisition periods. Our dependent variable is computed based on

two time windows. The first window captures the cumulative number of patents that an inventor

produced within the five years before the acquisition and the second window refers to the

cumulative number of patents in the five years after the acquisition.

Our explanatory variables capture the implementation of post-acquisition integration

decisions relying on information gathered by the DG research project and are defined as follows.

The variable R&D team reorganization is a dummy that takes value 1 if R&D teams of both the

acquiring and the acquired firms have been reorganized after acquisition, which imposes the

transfer of inventors across the two firms and organized them into new teams. The dummy

variable divestiture equals 1 if acquired R&D laboratories have been closed after the acquisition.

The variable R&D program diversification takes value 1 if R&D programs that are new relating

to the technological fields in which the acquired firm had developed distinctive capabilities prior

the acquisition are launched after the acquisition. Finally, the turnover of top managers

responsible for the acquired R&D function after the acquisition is captured by the dummy

 

22

 

acquired R&D top manager replaced. All these four variables are measured at the deal level and

subsequently shared by inventors working in the same firm.

We control for a number of acquisition- and individual-level characteristics. The first set

of controls concerns acquisition-specific characteristics. We control for the acquired firm’s

industry by including the dummy variable high tech which equals 1 for high-tech industry and 0

otherwise. This variable absorbs part of the cross-industry differences that could explain

differences in inventors’ propensity to patent. We also control for cross-border acquisitions by

using a dummy variable that indicates whether the acquiring and acquired firms are from

different countries. The variable hostile takes value 1 if the acquisition is the product of a hostile

takeover and 0 otherwise. Differently motivated acquisitions can implement more or less

invasive integration processes. During the DG research project, interviewed people were asked to

assess the importance of the following innovation related motives, on a five-point Likert scale

ranging from “not important at all” to “very important”: (1) economies of scale in R&D; (2)

economies of scope in R&D; (3) restructuration of the R&D process, elimination of duplicative;

(4) obtain access to assets, skills and capabilities of technological nature of the merging

companies; (5) obtain access to knowledge and other technical resources embedded in the

environment of the merging companies; (6) spread the risks of the R&D portfolio; (7) reduce the

risk of being imitated. Based on this information, we compute the variable innovation related

motives as a single composite measure based on the loadings from a principal component factor

analysis4 of the seven indicators of innovation related motives (Cronbach’s alpha = 0.95). The

existence of previous interactions or relationships between the acquired and the acquiring firm

should supposedly lower the uncertainties among R&D personnel in the acquired firm.

                                                            4 Factor loadings: item 1 = 0.97; item 2 = 0.97; item 3 = 0.97; item 4 = 0.94; item 5 = 0.69; item 6 = 0.97; item 7 =

0.96; Eigenvalue = 6.04; variance explained = 0.87 %. 

 

23

 

Therefore, we control for the variable technological links that equals 1 if acquiring and acquired

firms have experienced one or more technological alliances (i.e., equity joint ventures, license

agreements, minority shareholding) with one another before the focal acquisition. We also

control for the technological similarity that refers to the extent to which acquiring and acquired

firms operate in the same technological fields. The variable technology similarity is defined

using the index proposed by Makri, Hitt and Lane (2010)5. Using a five-year window regarding

the pre-acquisition period, we compute the ratio between the total number of patents filed by the

acquired and acquiring firms within the same 3-digit patent classes multiplied by the total

number of patents accumulated by the acquirer in all common classes, and the acquirer total

patents within this period. We also control for the relative size of the acquired firm using the

ratio between the total sales of the acquired and acquiring firm. Finally, we added year dummies

regarding the year in which the acquisition was concluded to control for differences in patenting

inventors’ behaviour originating from year trends.

The second set of controls concerns individual-specific characteristics. Specifically, we

control for the variable inventor tenure that is the number of years between the date of the first

patent that the inventor produced with the acquired firm as the assignee and the acquisition date.

We also control for the inventors search depth (Katila and Ahuja, 2002) in the five years prior to

the acquisition by computing the ratio between repeated citations and the total citations observed

in their patenting output. Finally, we control for the familiarity of the acquired inventor with the

acquirer’s expertise. Similar to a measure proposed by Parachuri et al. (2006) we first identify all

patents that had been produced by the acquirer firm within the five years before the acquisition.

Then, we observe the main class related to each of those patents in order to identify the field in

                                                            5 Our measure differs from Marki et al. (2010) in terms of the time widow used to compute it (i.e., five years), their

original measure was based on a single year and we consider the patents accumulated over five years

 

24

 

which the acquirer has been mostly active prior to the acquisition. Subsequently, we look at the

patents that each inventor in the acquired firm has produced and compute the ratio between the

inventor’s patents concentrated within the acquirer’s main technological area and the total

patents produced within the same period.

Coarsened exact matching

We use the CEM procedure to create the treatment and control sample with balanced

characteristics in terms of several pre-treatment covariates (Iacus et al., 2011; 2012). This

matching technique works through a process of data pruning in which the CEM algorithm creates

a set of strata that must contain at least one treatment and control observation, which allows to

subsequently running the analysis in the original (pruned) data (Blackwell et al., 2009). The

outcome of this matching process ensures a reduction in the overall imbalance between treated

and control groups in terms of the matching characteristics while relaxing several assumptions

that are required to produce unbiased estimates of the treatment effects (Iacus et al., 2011; 2012;

Aggarwal and Hsu, 2014).

To run the CEM algorithm we first split the 3,978 acquired inventors in terms of

treatment and control groups. It is interesting to note that in the case of 15 merged firms—2,083

(52.36%) inventors—the R&D team reorganization is implemented after acquisition while it

does not occur in 10 merged firms—1,895 (47.64%) inventors. In order to prune the data, we

match inventors in the treatment and control groups using the following five pre-treatment

observables pertaining to inventor-level characteristics, which we expect to be correlated with

the treatment. The first variable is inventor’s technological diversity which measures the

heterogeneity of the knowledge inventors have accumulated in different technological fields

within the five year prior to the acquisition. It is computed based on a Herfindahl index (HI)

 

25

 

calculated using the main technological fields (3-digit class) to which the inventor patents are

assigned (Hoisl, 2007). Second, because the status of an inventor before the acquisition can also

affect his post-deal performance, we also match both groups using a dummy variable indicating

if an individual is a star inventor. We identify as start inventors those individuals for which the

pre-acquisition patenting productivity is one standard deviation above the mean of the other

inventors observed within the same firm. The third variable is inventor’s last patenting

calculated as the number of years passed since the last time the focal inventor patented before the

acquisition (Paruchuri et al., 2006). Fourth, to ensure that the treatment and control group are

similar in terms of their performance in the pre-acquisition period, we also consider the variable

number of pre-patents which is the total number of patents each inventor produced within the

five years prior to the year that his/her firm was acquired. This last variable increases the

confidence that the deviations in the trajectory between individuals in the treatment and control

group are likely to be due to the treatment, as their ex-ante performance should be comparable.

Finally, we use three period dummies to match the inventors in the treatment and control groups

around the same years that their firms were acquired.

The CEM matching produced a significant improvement of the imbalance in our data,

with the overall imbalance provided by the L1 statistics moving from L1=0.33 to L1= 0.11 (for a

detailed discussion on the interpretation of the L1 statistic, see Iacus et al., 2011). The

improvement in the overall level of imbalance between the two groups is the most important

statistics of the CEM outcome, because although the marginal distribution of the five variables

used in the matching procedure could be similar their joint distribution could still be imbalanced

(Blackwell et al., 2009). During the matching process, 285 inventors (approximately 7% of the

original sample) were removed from the sample because the CEM algorithm was not able to

 

26

 

match those observations in a strata including at least one inventor from the treatment and one

for the control group. By dropping these dissimilar observations between inventors that

experienced the treatment and the ones that did not, CEM reduced the sample to 3,693

observations that will serve as the sample in which we will test our hypotheses. Among those

observations, 1,857 belong to the treatment group and 1,836 to the control group. Given that we

are not using a one-to-one matching solution we employ CEM weights to compensate for the

differential strata size (Blackwell et al., 2009).

A difference-in-differences approach for examining the effect of R&D team reorganization

The difference-in-differences approach exploits the fact that we observe the productivity of

inventors not just in the post-acquisition but also in the period that precedes it. Although the

post-acquisition differences in the treatment and control groups confound inherent differences

between the two groups, we can partially disentangle these effects by reducing the imbalance

between the two groups based on “observables” and tracking them in the pre- and post-

acquisition periods. Furthermore, although we make both groups comparable in terms of their

performance before the acquisition, we still look at pre-acquisition differences in patenting

performance for both groups and use it as a benchmark to examine the post-acquisition

differences and identify the part of the second difference that can be attributed to the

reorganization of teams itself. We use the following equation to implement this logic:

Patenting activityj,t = f (辟眺treatment group j + 辟眺牒treatment group j × post-acquisitionj,t +

辟牒 post-acquisitionj,t + 辟掴 Xj + įt – pre-acquisition (j) + ej,t)

For each inventor j whose R&D teams are reorganized after acquisition, the dummy variable

treatment group takes value 1 and value 0 if the inventors belong to the control group. In this

model, 辟眺captures the systematic differences between the treatment and control groups that

exist before the acquisition. The interaction term treatment × post-acquisition should capture the

 

27

 

net effect (net of the average acquisition effect) that the treatment has on the treated group.

Based on our theoretical argumentation, the coefficient of interest 辟眺牒, should be negative and

significant. The final variable post acquisition takes value 1 for both the treatment and control

groups only when observed in the post-acquisition period. Therefore, 辟牒 consists of the

counterfactual in the patenting productivity for the post-acquisition period in the case R&D team

reorganization had not happened.

Following previous studies (e.g., Meyer, 1995; Younge et al., 2014), we test our

hypothesized moderating effects by extending the basic difference-in-differences model using

three-way interactions between the term treatment × post-acquisition with each variable of

interest.

Results

Table 1 reports the means, standard deviations, minimum, maximum and Pearson correlation

coefficients for all variables used in the study, based on the CEM-matched sample of 3,693

unique inventors observed across two period (pre- and post-acquisition) and forming a total of

7,386 observations. We observe that the variable relative size is correlated with some of our

variables of interest. To test if those correlations are biasing our results, we tested the stability of

the coefficients by removing relative size from the models and the results remained the same.

[Insert Table 1 here]

Apart from the dependent variable and the variable post-acquisition the others variables are time

invariant, and vary within the sample only when interacted with the post-acquisition variable.

We start our analysis with a simple difference-in-differences estimation based on the

CEM sample and including no further explanatory variables. This estimation presents the basic

intuition behind our idea to use the pre- and post-acquisition performance to measure the effect

 

28

 

that the reorganization of R&D teams has on inventors patenting activity. Table 2 reports the

means for the dependent variable Patenting performance based on four subsamples: R&D team

reorganization (treatment) pre-acquisition, R&D team reorganization (treatment) post-

acquisition, control group pre-acquisition and control group post-acquisition. Based on Table 2,

we observe that in the pre-acquisition period the patenting performance of inventors in the

treatment (3.039) and control group (2.954) is not statistically significant. When we consider the

relative changes from the pre- to the post-acquisition period, it is possible to observe that the

inventors of both groups had a substantial decline in their patenting productivity. Indeed, the

inventors in the control group presented an average decline in the order of 46% (i.e., 1 −

(1.586/2.954)) in the post-acquisition period, while for the treatment group this decline was in

the order of 69% (i.e., 1 − (0.913/3.039)). This finding is in line with previous studies that have

reported a negative average effect of acquisitions on the acquired inventors. Furthermore, it also

confirms the suitability of using as a control sample a group of inventors that also went through

an acquisition process in order to have a more conservative counterfactual for the effect of R&D

team reorganization on patenting performance.

[Insert Table 2 here]

The post-acquisition mean values for the patenting productivity in the treatment and control

inventors indicate a statistically significant difference (p<0.01) between the two groups, with the

control group reporting a mean value of 1.212 patents while the treatment 0.913 patents. If we

subtract from the mean value observed for the treatment the counterfactual provided by the

control group we observe a second difference in the order of −0.673 (0.913 − 1.212). Given that

the in the pre-acquisition period both groups are not statistically different, the mean comparison

of the post-acquisition values would provide an indication of the negative effect that team

 

29

 

reorganization has on inventor patenting productivity. Nevertheless, following the difference-in-

difference logic, we proceed and obtain the final estimates by subtracting the second difference

from the first (−0.673 − 0.085 = −0.758). Those estimates indicate that when we subtract the pre-

acquisition differences from the second difference, the remaining value is negative and

statistically significant (p<0.01).

To test our hypotheses, we estimate a negative binomial model using a difference-in-

differences setup (see Table 3). The variable post-acquisition × R&D team reorganization is

used to estimate the treatment on the treated observations; around 25% of the observations across

the two periods have a positive outcome for this variable. Our analyses are conducted at the

inventor-period level. As such, each inventor is observed and recorded both in the pre- and the

post-acquisition period. We model unobserved heterogeneity at the inventor level with random

effects. Finally, while we match the treatment and control groups on inventor level

characteristics, the additional set of control variables used in the main analysis helps removing

undesirable residual heterogeneity that could remain in the sample after the matching process.

We include an indicator for the treatment group R&D team reorganization, and indicator for

post-acquisition and the interaction capturing the effect of the treatment on treated post-

acquisition × R&D team reorganization. Model 1 reports the controls and the variables of

interest before the interaction testing their effect in the post-acquisition period. The effect for the

variable post-acquisition is negative and highly significant (p<0.001), this result is closely

aligned with previous studies. Model 2 introduces the interaction between post-acquisition ×

R&D team reorganization. In line with Hypothesis 1 this effect is negative and significant

(p<0.001). Model 3 introduces the set of interaction terms testing the moderating effect of

divesture on R&D team reorganization, the results for the three-way interaction divestiture ×

 

30

 

post-acquisition × R&D team reorganization is negative and significant (p<0.001), confirming

Hypothesis 2. Model 4 tests the moderating effect for R&D program diversification. As we

predicted by Hypothesis 3, the coefficient of the interaction R&D program diversification ×

post-acquisition × R&D team reorganization is also negative and significant (p<0.001). Finally,

testing Hypothesis 4, Model 5 enters the moderating effect acquired R&D top management

replaced. In line with our expectations, the results for the term acquired R&D top management

replaced × post-acquisition × R&D team reorganization is positive and significant (p<0.001).

To test the stability of the coefficients in Model 6 we enter the main effect together with the three

moderating coefficients. Our findings are largely confirmed with the only exception of the

moderating effect of technological integration through the launch of R&D programs in

technological fields new to the acquired firm which effect becomes insignificant.

Robustness checks, available from authors upon request, are not shown here because of

space limitation. First, in order to test if our results are biased by omitted time-invariant

individual level characteristics we estimate the same models with inventor level fixed effects.

The results of a negative binomial with fixed effects are similar to the ones reported above.

Second, as potential unobserved heterogeneity at the deal (acquisition) level could also be a

source of bias, we estimated the models using deal dummies. Those dummy variables absorb

time-invariant characteristics that relate to the acquisition itself. Those estimations were also

equivalent to the reported negative binomial model with random effects. The fixed effects

estimators were not our preferred choice because they do not allow time invariant variables to be

included in the models. However, given that our main results remain qualitatively unchanged, we

found no evidence of omitted variables biasing our estimates.

DISCUSSION AND CONCLUSIONS

 

31

 

Discrete intangible and tacit resources are difficult to exchange in the market due to market

failures. Accordingly, the strategic literature often proposes horizontal acquisitions as a choice to

access such resources that otherwise are not easily transferable (e.g., Bowman and Singh, 1993;

Karim and Mitchell, 2000). However, reaping the synergistic benefit of the strategy behind these

acquisitions requires that the resources of both acquiring and acquired firms are reconfigured,

realigned, and rationalized, that is integration decisions are implemented (Capron, 1999; Karim

and Mitchell, 2000; Cording et al., 2008).

In relation to the effect of horizontal acquisitions on innovative performance, research

shows that the post-acquisition integration, although necessary to realize synergistic benefits,

imposes several challenges on the inventive labor force of the acquiring and acquired firms

which, ultimately, can harm innovation (Ernst and Vitt, 2000; Ranft and Lord, 2000; Ranft and

Lord, 2002; Puranam et al., 2006). In particular, when acquired firms are formally integrated

under the corporate umbrella of the acquiring firms (structural integration), horizontal

acquisitions harm knowledge-workers’ motivation and create disruption, expose firms’ R&D

personnel to feelings of resentment and dissatisfaction which result in reduced inventor

productivity (Paruchuri et al., 2006). We contribute to this work about the relationship between

integration and inventors’ innovation performance, addressing the complexity of post-acquisition

integration and going beyond structural integration and the categorization of integration with

single types of decisions. Specifically, to better disentangle the effect of the complex post-

acquisition integration process on inventor productivity, we focus on the decision of reorganizing

R&D teams and its interdependencies with three additional distinct integration decisions that

capture physical, technological, and managerial integration after acquisition.

 

32

 

Our results help illuminating the internal dynamics by which integration after horizontal

acquisitions affects innovation by extending the understanding of how integration decisions

hinge on inventors’ choices and actions. We discussed how integration through R&D team

reorganization directly creates the conditions for higher coordination, common language and

knowledge sharing between inventors of the acquiring and the acquired firms. However, to the

extent that team reorganization implies a change in team routines it can also harm knowledge

recombination. In line with the idea that the characteristics of tacitness, inimitability and firm-

specificity of organizational knowledge and routines create inertia and rigidity which inhibit firm

organizational change (Leonard-Barton, 1992), we find that team reorganization imposing

changes and adaptation of team-specific routines inhibits team learning and negatively affect

team members’ productivity. However, the effect of team reorganization on inventor

productivity is not independent from other integration decisions firms might implement

concurrently. Our findings support this expectation. We find that the closure of acquired R&D

laboratories and the launch of R&D programs in technological field new to the acquired firms

aggravate the negative effect of R&D team reorganization on inventor productivity. On the

contrary, the negative effect of team reorganization is alleviated when the acquired R&D top

manager is replaced.

To fully understanding how value is created from acquisition, post-acquisition research

needs to take into account the multifaceted nature of integration (Haspeslagh and Jemison,

1991). Our findings strongly support this view by theoretically arguing and empirically showing

how the effects of different integration decisions acting upon the interdependence between

resources of the acquired and acquiring firms are intertwined. Specifically, the divestiture of

redundant assets which is often a necessary decision to implement physical integration

 

33

 

(Shrivastava, 1986; Capron, 1999; Capron et al., 2001; Anand, 2004), adds additional stress and

sense of disruption to inventors already engaged in higher efforts to learn and align their routines

with the dominant (or new) routines of the team in which they have been reorganized. Also, in

line with team literature showing that successful team members will have to well-perform task-

related as well as team-related functions (Mcintyre and Salas, 1995; Mathieu et al., 2000), we

highlight that post-acquisition integration decisions which have an impact on task- and team-

related routines moderate the negative effect of team reorganization of inventors’ productivity.

First, we looked at technological integration which creating the conditions to specialize or

broaden the firm’s technological competencies affects task-related routines. Acquired inventors

had developed skills and routines as a result of their work experience within the technological

fields in which the acquired firm is specialized. Team reorganization—placing together acquired

and acquiring inventors with different information, knowledge, and expertise—can potentially

promote search behaviors that lead to innovative solutions. Firms can enhance this process by

formally launching new R&D programs. However, this decision exacerbates the effect of

integration through team reorganization on innovative performance by imposing inventors to

learn new task-related routines. Second, we examined managerial integration and found that the

replacement of the acquired R&D top manager facilitates the homogenization of team-related

routines by forcing unlearning of old and different team-related routines. Highly similar team-

related routines result in more effective team processes that alleviate the main negative effect of

integration on team members’ performance.

This final result has also broader implications for the literature on acquisition and top

management turnover. Management scholars have suggested that retaining acquired R&D

managers results in better coordination and knowledge integration after acquisition (Graebner,

 

34

 

2004). However, parallel research also shows that the replacement of the acquired R&D top

manager will be beneficial for R&D performance when the similarity between acquiring and

acquired firms increases (Colombo and Rabbiosi, 2014). We add to this work on the role of

acquired R&D top managers showing the importance of qualifying the effect of one integration

decision in relation to other concurrent integration decisions. On one hand, we find that replacing

acquired R&D top manager is beneficial; on the other hand, we also confirm that the replacement

per se as a negative effect on the inventor innovative performance after acquisition.6 In line with

our theoretical arguments, by replacing acquired R&D top managers firms can enhance

unlearning processes which favor the adoption of new team-related routines and, therefore,

reduce the negative effects on inventor productivity of team reorganization that directly induces

changes on inventors’ team routines. However, the replacement of the acquired R&D top

manager can be detrimental if the team work environment of acquired inventors is not

reorganized.

                                                            6 In table 3, Models 5and 6 show a negative and significant coefficient of the variable acquired R&D top

management replaced × post-acquisition.

 

35

 

REFERENCES

Aggarwal VA, Hsu DH. 2014. Entrepreneurial exits and innovation. Management Science 60:

867-887.

Ahuja G, Katila R. 2001. Technological acquisitions and the innovation performance of

acquiring firms: A longitudinal study. Strategic Management Journal 22: 197-220.

Alderfer CP. 1987. An intergroup perspective on organizational behavior. In J Lorsch (Ed.),

Handbook of Organizational Behavior: 190-222. Prentice-Hall: Englewood Cliffs, NJ.

Anand J. 2004. Redeployment of Corporate Resources: A Study of Acquisition Strategies in the

US Defense Industries, 1978-1996. Managerial and Decision Economics 5(6/7): 383-400.

Argote L, Gruenfeld D, Naquin C. 2001. Group learning in organizations. In ME Turner (Ed.),

Groups at work: Theory and research: 369-411. LEA: Mahwah, NJ.

Arrow H, McGrath JE, Berdahl JL. 2000. Small groups as complex systems: Formation

coordination, development and adaptation. Sage: Thousand Oaks, CA.

Becker MC. 2004. Organizaional routines: a review of the literature. Industrial and Corporate

Change 13(4): 643-677.

Blackwell M, Iacus S, King G, Porro G. 2009. CEM: Coarsened exact matching in Stata. Stata

Journal 9(4): 524-546.

Bowman EH, Singh H. 1993. Corporate restructuring: reconfiguring the firm. Strategic

Management Journal 14: 5-14.

Cannon-Bowers JA, Salas E, Converse SA. 1993. Shared mental models in expert team decision

making. In NJ Castellan Jr. (Ed.), Current issues in individual and group decision making:

221-246. Erlbaum: Hillsdale, NJ.

Capron L. 1999. The long-term performance of horizontal acquisitions. Strategic Management

Journal 20: 987-1018.

Capron L, Dussauge P, Mitchell W. 1998. Resource redeployment following horizontal

acquisitions in Europe and North America, 1988-1992. Strategic Management Journal

19(7): 631-661.

Capron L, Hulland J. 1999. Redeployment of brands, sales forces, and general marketing

management expertise following horizontal acquisitions: A resource-based view. Journal of

Marketing 63(April): 41-54.

Capron L, Mitchell W, Swaminathan A. 2001. Asset Divestiture Following Horizontal

Acquisitions: A dynamic view. Strategic Management Journal 22: 817–844.

Cassiman B, Colombo MG, Garrone P, Veugelers R. 2005. The impact of M&A on the R&D

process - An empirical analysis of the role of technological- and market-relatedness.

Research Policy 34(2): 195-220.

Cohen WM, Levinthal DA. 1990. Absorptive Capacity: A New Perspective of Learning and

Innovation. Administrative Science Quarterly 35: 128-152.

Colombo MG, Rabbiosi L. 2014. Technological similarity, post-acquisition R&D reorganization

and innovation performance in horizontal acquisition. Research Policy 43: 1039-1054.

Cording M, Christmann P, King DR. 2008. Reducing causal ambiguity in acquisition integration:

intermediate goals as mediators of integration decisions and acquisitions performance.

Academy of Management Journal 51(4): 744-767.

Cyert RM, March JG. 1963. A behavioral theory of the firm. Englewood Cliffs: NJ.

Datta DK. 1991. Organizational fit and acquisition performance: Effects of post-acquisitions

integration. Strategic Management Journal 12: 281-297.

 

36

 

Dierickx I, Cool K. 1989. Asset stock accumulation and sustainability of competitive advantage.

Management Science 35: 1504-1511.

Dosi G. 1982. Technological paradigms and technological trajectories. Research Policy 11: 147-

162.

Edmondson AC. 1999. Psychological Safety and Learning Behavior in Work Teams.

Administrative Science Quarterly 44: 350-383.

Edmondson AC. 2002. The local and variegated nature of learning in organizations.

Organization Science 13(2): 128-146.

Edmondson AC. 2003. Speaking Up in the Operating Room: How Team Leaders Promote

Learning in Interdisciplinary Action Teams. Journal of Management Studies 40(6): 1419-

1452.

Edmondson AC, Bohmer RM, Pisano GP. 2001. Disrupted routines: Team learning and new

technology implementation in hospitals. Administrative Science Quarterly 46(4): 685-716.

Edmondson AC, Nembhard IM. 2009. Product Development and Learning in Project Teams:

The Challenges Are the Benefits. Journal of Product Innovation Management 26: 123–138.

Ernst H, Vitt J. 2000. The influence of corporate acquisitions on the behavior of key inventors.

R&D Management 30(2): 105-119.

Felin T, Foss NJ. 2005. Strategic Organization: a Field in Search of Micro-Foundations.

Strategic Organization 3: 441-455.

Felin T, Hesterly W. 2007. The Knowledge-Based View, Nested Heterogeneity, and New Value

Creation: Philosophical Considerations on the Locus of Knowledge. Academy of

Management Review 32(1): 195-218.

Graebner ME. 2004. Momentum and serendipity: how acquired leaders create value in the

integration of technology firms. Strategic Management Journal 25: 751-777.

Grant RM. 1996. Toward a knowledge-based theory of the firm. Strategic Management Journal

17(Winter special issue): 109-122.

Hackman JR. 1987. The design of work teams. In J Lorsch (Ed.), Handbook of Organizational

Behavior: 315-342. Prentice-Hall: Englewood Cliffs, NJ.

Haleblian J, Devers CE, McNamara G, Carpenter MA, Davison RB. 2009. Taking Stock of What

We Know About Mergers and Acquisitions: A Review and Research Agenda. Journal of

Management 35(3): 469-502.

Hambrick DC, Cannella AA. 1993. Relative standing: a framework for understanding departures

of acquired executives Academy of Management Journal 36(4): 733-762.

Haspeslagh PC, Jemison DB. 1991. Managing Acquisitions. Free Press: New York.

Henderson RM, Clark KB. 1990. Architectural innovation: The reconfiguration of existing

product technologies and the failure of established firms. Administrative Science Quarterly

35(1): 9-30.

Hitt MA, Hoskisson RE, Ireland RD, Harrison JS. 1991. Effects of acquisitions on R&D inputs

and outputs. Academy of Management Journal 34(3): 693–706.

Hoisl K. 2007. Tracing mobile inventors—The causality between inventor mobility and inventor

productivity. Research Policy 36(5): 619-636.

Huff AS. 1982. Industry Influences on Strategy Formulation. Strategic Management Journal 3:

119-130.

Iacus SM, King G, Porro G. 2011. Multivariate Matching Methods That Are Monotonic

Imbalance Bounding. Journal of the American Statistical Association 106(493): 345-361.

 

37

 

Iacus SM, King G, Porro G. 2012. Causal Inference without Balance Checking: Coarsened Exact

Matching. Political Analysis 20(1): 1-24.

Jemison DB, Sitkin SB. 1986. Corporate Acquisitions: A Process Perspective. Academy of

Management Review 11(1): 145-163.

Kapoor R, Kwanghui LIM. 2007. The impact of acquisitions on the productivity of inventors at

semiconductor firms: a synthesis of knowledge-based and incentive-based perspectives.

Academy of Management Journal 50(5): 1133-1155.

Kapoor R, Lim K. 2007. The impact of acquisitions on the productivity of inventors at

semiconductor firms: A synthesis of knowledge-based and incentive-based perspectives. The

Academy of Management Journal 50(5): 1133-1155.

Karim S, Mitchell W. 2000. Path-dependent and path-breaking change: reconfiguring business

resources following acquisitions in the U.S. medical sectors, 1978-1995. Strategic

Management Journal 21: 1061-1081.

Katila R, Ahuja G. 2002. Something old, something new: a longitudinal study of search behavior

and new product introduction. Academy of Management Journal 45(6): 1183-1194.

Klein JA, Posey PA. 1986. Good Supervisors Are Good Supervisors--Anywhere. Harvard

Business Review 64(6): 125-128.

Klimoski R, Mohammed S. 1994. Team mental model: Construct or metaphor? Journal of

Management 20: 403-437.

Kogut B, Zander I. 1996. What firms do? Coordination, identity, and learning. Organization

Science 7: 502-510.

Kogut B, Zander U. 1992. Knowledge of the firm, combinative capabilities, and the replication

of technology. Organization Science 3: 383-397.

Kogut B, Zander U. 1993. Knowledge of the Firm and the Evolutionary-Theory of the

Multinational-Corporation. Journal of International Business Studies 24(4): 625-645.

Kraiger K, Wenzel LH. 1997. Conceptual development and empirical evaluation of measures of

shared mental models as indicators of team effectiveness. In MT Brannick, E Salas, C Prince

(Eds.), Team performance assessment and measurement: Theory, methods, and

applications: 62-84. Erlbaum: Mahwah, NJ.

Larsson R, Finkelstein S. 1999. Integrating Strategic, Organizational, and Human Resource

Perspectives on Mergers and Acquisitions: A Case Survey of Synergy Realization.

Organization Science 10(1): 1-26.

Leonard-Barton D. 1992. Core capabilities and core rigidities: A paradox in managing new

product development. Strategic Management Journal 13: 111-125.

Levin SG, Stephan PE. 1991. Research productivity over the life cycle: Evidence for academic

inventors. American Economic Review 81: 114-132.

Levine JM, Richard M, Moreland L. 1999. Knowledge transmission in work groups: Helping

newcomers to succeed. In LL Thompson, JM Levine, DM Messick (Eds.), Shared cognition

in organizations: The management of knowledge: 267-296. Lawrence Erlbaum: Mahwah,

NJ.

Lewis K, Lange D, Gillis L. 2005. Transactive memory systems, learning, and learning transfer.

Organization Science 16(6): 581-598.

Li GC, Lai R, D’Amour A, Doolin DM, Sun Y, Torvik VI, Yu A, Fleming L. 2014.

Disambiguation and co-authorship networks of the US patent inventor database (1975–

2010). Research Policy 43(6): 941-955.

 

38

 

Makri M, Hitt M, Lane PJ. 2010. Complementary technologies, knowledge relatedness and

innovation outcomes in high-technology M&As. Strategic Management Journal 31(6): 602-

628.

Manz CC, Newstrom J. 1990. Self-managing teams in a paper mill: Success factors, problems,

and lessons learned. International Human Resource Management Review 1: 43-60.

Mathieu JE, Heffner TS, Goodwin GF, Salas E, Cannon-Bowers JA. 2000. The influence of

shared mental models on team process and performance. Journal of Applied Psychology

85(2): 273-283.

Mcintyre RM, Salas E. 1995. Measuring and managing for team performance: Emerging

principles from complex environments. In R Guzzo, E Salas (Eds.), Team effectiveness and

decision making in organizations: 149-203. Jossey-Bass: San Francisco.

Meyer BD. 1995. Natural and Quasi-Experiments in Economics. Journal of Business &

Economic Statistics 13(2): 151-161.

Mohammed S, Ferzandi L, Hamilton K. 2010. Metaphor No More: A 15-Year Review of the

Team Mental Model Construct. Journal of Management 36(4): 876-910.

Mudrack PE. 1989. Defining group cohesiveness: A legacy of confusion? Small Group Behavior

20: 37-49.

Napier NK. 1989. Mergers and acquisitions, human resource issues and outcomes: A review and

suggested typology. Journal of Management Studies 26: 271-289.

Nelson R, Winter S. 1982. An evolutionary theory of economic change. Harvard University

Press: Cambridge, MA.

Nystrom PC, Starbuck WH. 1984. To avoid organizational crises, unlearn. Organizational

Dynamics 12(4): 53-65.

Orlikowski WJ. 1993. Learning from notes: Organizational issues in groupware implementation.

Information Society 9: 237-250.

Pablo AL. 1994. Determinants of acquisition integration level: A decision making perspective.

Academy of Management Journal 37: 803-836.

Paruchuri S, Eisenman M. 2012. Microfoundations of Firm R&D Capabilities: A Study of

Inventor Networks in a Merger. Journal of Management Studies 49(8): 1509-1535.

Paruchuri S, Nerkar A, Hambrick DC. 2006. Acquisition integration and productivity losses in

the technical core: disruption of inventors in acquired companies. Organization Science

17(5): 545-562.

Puranam P, Singh H, Zollo H. 2006. Organizing for Innovation: Managing the Coordination-

Autonomy Dilemma in Technology Acquisitions. The Academy of Management Journal

49(2): 263-280.

Ranft AL, Lord MD. 2000. Acquiring new knowledge: The role of retaining human capital in

acquisitions of high-tech firms. The Journal of High Technology Management Research

11(2): 295-319.

Ranft AL, Lord MD. 2002. Acquiring New Technologies and Capabilities: A Grounded Model

of Acquisition Implementation. Organization Studies 13(4): 420–441.

Rentsch JR, Klimoski R. 2001. Why do ‘great minds’ think alike?: Antecedents of team member

schema agreement. Journal of Organizational Behavior 22: 107-120.

Rico R, Sánchez-Manzanares M, Gil F, Gibson C. 2008. Team Implicit Coordination Processes:

A Team Knowledge-Based Approach. The Academy of Management Review 33(1): 163-184.

Senge P. 1990. The Fifth Discipline: The Art and Practice of the Learning Organization.

Currency Doubleday: New York.

 

39

 

Shrivastava P. 1986. Postmerger integration. Journal of Business Strategy 7(1): 65-76.

Singh J, Agrawal A. 2011. Recruiting for Ideas: How Firms Exploit the Prior Inventions of New

Hires. Management Science 57(1): 129-150.

Stasser G, Stewart DD, Wittenbaum GM. 1995. Expert roles and information exchange during

discussion: The importance of knowing who knows what. Journal of Experimental Social

Psychology 31(3): 244-265.

Stewart GL, Barrick MR. 2000. Team structure and performance: Assessing the mediating role

of intrateam process and the moderating role of task type. Academy of Management Journal

43(2): 135-148.

Szulanski G. 1996. Exploring internal stickiness: impediments to the transfer of best practice

within the firm. Strategic Management Journal 17: 27-43.

Tushman ML, Romanelli E. 1985. Organizational evolution: A metamorphosis model of

convergence and reorientation. In LL Cummings, BM Staw (Eds.), Research in

organizational behavior, Vol. 7: 171-222. JAI Press: Greenwich, CT.

Walsh JP. 1989. Doing a deal: Merger and acquisition negotiations and their impact upon target

company top management turnover. Strategic Management Journal 10(4): 307-322.

Wiersema MF, Bantel KA. 1993. Top management team turnover as an adaptation mechanism:

The role of the environment. Strategic Management Journal 14: 485-504.

Younge KA, Tong TW, Fleming L. 2014. How anticipated employee mobility affects acquisition

likelihood: Evidence from a natural experiment. Strategic Management Journal: n/a-n/a.

Zollo M, Singh H. 2004. Deliberate learning in corporate acquisitions: Post-acquisition strategies

and integration capability in U.S. bank mergers. Strategic Management Journal 25: 1233-

1256.

 

40

 

TABLES

Table 1. Descriptive Statistics and Correlations Coefficients (N = 7,386; Units= 3,693)

Variable Mean S.D. Min Max [1] [2] [3] [4] [5] [6] [7] [8] [9]

[1] Patenting performance 1.75 3.1 0.00 43.00 1.00

[2] Post-acquisition x R&D team reorganization 0.25 0.43 0.00 1.00 -0.16 1.00

[3] Post-acquisition 0.50 0.5 0.00 1.00 -0.27 0.58 1.00

[4] R&D team reorganization 0.50 0.5 0.00 1.00 0.07 0.58 0.00 1.00

[5] Divestiture 0.45 0.5 0.00 1.00 0.09 0.36 0.00 0.62 1.00

[6] R&D program diversification 0.46 0.5 0.00 1.00 -0.03 0.28 0.00 0.49 0.73 1.00

[7] Acquired R&D top manager replaced 0.15 0.36 0.00 1.00 0.10 0.09 0.00 0.16 -0.05 -0.20 1.00

[8] High tech 0.08 0.27 0.00 1.00 0.11 -0.05 0.00 -0.08 -0.26 -0.06 0.15 1.00

[9] Cross-border 0.89 0.31 0.00 1.00 -0.06 0.13 0.00 0.22 0.31 0.03 -0.03 -0.62 1.00

[10] Hostile 0.69 0.46 0.00 1.00 -0.07 -0.05 0.00 -0.09 0.02 0.16 -0.58 0.05 -0.08

[11] Innovation related motives 0.02 1.03 -0.37 3.29 0.11 -0.22 0.00 -0.38 0.02 0.04 -0.16 0.57 -0.50

[12] Technological links 0.07 0.25 0.00 1.00 -0.05 -0.07 0.00 -0.12 -0.25 -0.10 0.62 0.20 0.09

[13] Technology similarity 0.62 0.17 0.00 0.91 0.08 -0.28 0.00 -0.49 -0.15 -0.43 -0.31 0.14 0.09

[14] Relative size 148.75 93.08 0.00 280.55 0.00 0.41 0.00 0.72 0.74 0.71 -0.33 -0.12 0.31

[15] Inventor tenure 5.22 5.67 0.00 24.00 0.15 0.10 0.00 0.18 0.23 0.16 0.17 0.00 0.03

[16] Search depth 0.14 0.28 0.00 1.00 0.33 0.05 0.00 0.08 0.02 0.03 0.01 0.20 -0.08

[17] Familiarity with acquirer’s expertise 0.09 0.27 0.00 1.00 0.06 -0.13 0.00 -0.22 -0.18 -0.16 0.12 0.24 -0.18

Variable Mean S.D. Min Max [10] [11] [12] [13] [14] [15] [16] [17]

[10] Hostile 0.69 0.46 0.00 1.00 1.00

[11] Innovation related motives 0.02 1.03 -0.37 3.29 -0.18 1.00

[12] Technological links 0.07 0.25 0.00 1.00 -0.41 -0.09 1.00

[13] Technology similarity 0.62 0.17 0.00 0.91 0.47 0.26 -0.30 1.00

[14] Relative size 148.75 93.08 0.00 280.55 0.40 -0.21 -0.20 -0.12 1.00

[15] Inventor tenure 5.22 5.67 0.00 24.00 -0.05 -0.03 0.15 -0.09 0.17 1.00

[16] Search depth 0.14 0.28 0.00 1.00 -0.02 0.13 0.03 0.04 0.08 0.12 1.00

[17] Familiarity with acquirer’s expertise 0.09 0.27 0.00 1.00 -0.05 0.25 0.26 0.10 -0.16 0.03 0.05 1.00

 

 

 

 

 

41

 

Table 2. Differences in Inventors Patenting Productivity for the pre- and post-acquisition periods 

Average Patent Count at the Inventor level

Pre-acquisition Post-acquisition

Treatment Group Subsample mean: Subsample mean: First difference (row):

(Acquisition with R&D team

Reorganization) Patent Count= 3.039 Patent Count= 0.913 Patent Count= -2.126

(N= 1,857) (0.086) (N= 1,857) (0.086) (N= 3,714)

Control Group Subsample mean: Subsample mean: First difference (row):

(Acquisition with no R&D

team Reorganization) Patent Count= 2.954 Patent Count= 1.586 Patent Count= -1.368

(N=1,836) (0.087) (N=1,836) (0.087) (N= 4,136)

Differences First difference (Column) First difference (Column) Difference-in-differences:

Patent Count= 0.085 Patent Count= -0.673*** Patent Count= -0.758***

(N= 3,693) (0.123) (N= 3,693) (0.123) (0.173)

*** p<0.01; ** p<0.05; * p<0.10

 

 

42

 

Table 3. Panel Negative Binomial Regression Models with Random Effects

Variables Model 1 Model 2 Model 3 Model 4 Model 5 Model 6

Post-acquisition × R&D team reorganization -0.635*** -0.331** -0.143† -0.975*** -0.528***

(0.061) (0.103) (0.079) (0.072) (0.123)

Divestiture × Post-acquisition × R&D team reorganization -0.628*** -1.028***

(0.145) (0.202)

Divestiture × R&D team reorganization -0.490 0.784

(0.564) (1.041)

Divestiture × Post-acquisition 0.302** 0.937***

(0.098) (0.167)

R&D program diversification × Post-acquisition × R&D

team reorganization -0.779*** 0.261

(0.128) (0.192)

R&D program diversification × R&D team reorganization -1.075 -1.548

(0.679) (1.306)

R&D program diversification × Post-acquisition -0.096 -0.794***

(0.089) (0.153)

Acquired R&D top management replaced × Post-acquisition

× R&D team reorganization 1.613*** 1.342***

(0.200) (0.212)

Acquired R&D top management replaced × R&D team

reorganization 2.275 0.299

(1.606) (0.529)

Acquired R&D top management replaced × Post-acquisition -0.699*** -0.740***

(0.176) (0.178)

Post-acquisition -1.389*** -1.089*** -1.142*** -1.058*** -1.038*** -0.997***

(0.031) (0.039) (0.043) (0.044) (0.039) (0.045)

R&D team reorganization -0.173 -0.033 0.134 0.425 0.442 0.507

(0.311) (0.312) (0.371) (0.418) (0.381) (0.420)

Divestiture 0.384* 0.375* 0.939 0.219 -0.151 -0.779

(0.165) (0.165) (0.623) (0.198) (0.370) (1.233)

R&D program diversification -0.249 -0.238 -0.682 0.247 0.178 1.028

(0.201) (0.202) (0.462) (0.337) (0.340) (1.218)

Acquired R&D top manager replaced 0.349 0.294 0.721 0.908* -2.758 0.119

(0.327) (0.328) (0.496) (0.462) (2.058) (0.517)

High tech 0.910† 0.836 1.339† 1.096† -3.502 -0.277

(0.543) (0.545) (0.714) (0.562) (2.776) (0.499)

Cross-border 0.168 0.109 0.327 0.700 -2.094 0.166

(0.377) (0.380) (0.424) (0.500) (1.459) (0.631)

 

43

 

Hostile -0.868*** -0.834*** -0.847*** -0.196 0.739 0.299

(0.234) (0.234) (0.231) (0.437) (1.011) (0.654)

Innovation related motives -0.028 -0.032 0.112 0.187 -0.157 0.077

(0.130) (0.131) (0.181) (0.168) (0.159) (0.143)

Technological links -1.235* -1.137* -1.755* -1.324* 3.932 0.301

(0.570) (0.573) (0.789) (0.579) (3.263) (0.624)

Technology similarity 0.585 0.657 -0.280 -0.219 2.911† 0.948

(0.667) (0.671) (1.088) (0.841) (1.596) (0.765)

Relative size 0.004* 0.004† 0.007* 0.006** -0.007 0.001

(0.002) (0.002) (0.004) (0.002) (0.007) (0.003)

Inventor tenure 0.019*** 0.019*** 0.020*** 0.020*** 0.020*** 0.020***

(0.002) (0.002) (0.002) (0.002) (0.002) (0.002)

Search depth 1.200*** 1.222*** 1.223*** 1.239*** 1.237*** 1.248***

(0.044) (0.044) (0.044) (0.044) (0.044) (0.044)

Familiarity with acquirer’s expertise 0.248*** 0.261*** 0.263*** 0.272*** 0.273*** 0.275***

(0.057) (0.056) (0.057) (0.057) (0.057) (0.057)

Year Dummies YES YES YES YES YES YES

Constant -0.808 -0.735 -2.061 -3.002† 5.175 -0.920

(1.075) (1.080) (1.632) (1.626) (3.949) (1.212)

Ln (r) Constant 2.138*** 2.160*** 2.156*** 2.168*** 2.170*** 2.176***

(0.058) (0.057) (0.057) (0.056) (0.056) (0.056)

Ln (s) Constant 2.134*** 2.102*** 2.098*** 2.064*** 2.061*** 2.036***

(0.074) (0.073) (0.073) (0.071) (0.071) (0.070)

Number of Observations 7.386 7.386 7.386 7.386 7.386 7.386

Chi2 4055.543*** 4097.047*** 4100.084*** 4051.216*** 4002.890*** 4019.368***

Log Likelihood -12015.078 -11959.832 -11950.258 -11914.703 -11905.702 -11874.298

Likelihood Ratio Comparison 110.492*** 129.640*** 200.750*** 218.752*** 281.560***

† p<0.10, * p<0.05, ** p<0.01, *** p<0.001